Interfacing the Biobase Biochemistry Analyzer with a LIS via HL7 over TCP/IP

1. Overview

The Biobase biochemistry analyzer communicates results over a local TCP/IP connection. The middleware listens on a fixed socket, receives a one-shot HL7 v2.3.1 ORU^R01 message encoded in hexadecimal, decodes and parses it, then writes structured result records into the Prolab LIS database — matching results to orders via barcode when present.

Stack:

• Transport: TCP socket — middleware listens on 192.168.1.96:502
• Protocol: HL7 v2.3.1 ORU^R01
• Encoding: Raw hex over TCP (single transmission per result set)
• Target LIS: Prolab

2. Message Flow

Biobase Analyzer
     │
     │  TCP connect → 192.168.1.96:502
     │  Sends: hex-encoded HL7 ORU^R01 (one shot)
     ▼
Linux Middleware (proline)
     │
     ├─ 1. Receive raw hex bytes
     ├─ 2. Decode hex → UTF-8 HL7 string
     ├─ 3. Parse HL7 segments (MSH / PID / OBR / OBX)
     ├─ 4. Extract barcode from PID-4 or OBR-3
     ├─ 5. Map OBX lines → Prolab result records
     └─ 6. Write to Prolab LIS

3. HL7 Message Structure

A typical transmission contains one MSH block per result set. The segments of interest are:

Example OBX line:

OBX|0|NM|246|AST BB|46.73|U/L|0~40|H|||||46.73|20260308160336||1000||

Fields: sequence | valueType | code | name | value | units | range | flag | ... | timestamp

4. Hex Decoding

The analyzer transmits the HL7 payload as a hex string. The first middleware step is decoding:

raw_hex = receive_from_socket()           # e.g. "4d53487c5e7e5c26..."
hl7_string = bytes.fromhex(raw_hex).decode("utf-8")
segments = hl7_string.strip().split("r")  # HL7 segment delimiter is CR

Note: HL7 v2 uses carriage return (r, 0x0D) as the segment terminator, not newline.

5. Barcode Extraction

When a barcode is present, it appears in two locations:

• PID-4 — patient/sample barcode (e.g. 142 for patient Fatou BA)
• OBR-3 — filler order number / sample ID (e.g. 21346159068142)

Anonymous samples (bench QC, calibrators) have no barcode; the middleware falls back to the internal messageControlId for record matching.

pid_fields = segments["PID"].split("|")
barcode = pid_fields[3] if len(pid_fields) > 3 and pid_fields[3] else None

obr_fields = segments["OBR"].split("|")
sample_id = obr_fields[2]  # OBR-3: filler order number

6. OBX → Prolab Record Mapping

Each OBX segment maps to one result line in Prolab:

Flags are normalized before insertion: H → High, L → Low, N → Normal (or absent).

7. Two Message Patterns

The Biobase emits two distinct PID patterns depending on context:

Anonymous sample (no patient, bench test):

PID|17||||BS1|||O|...

Barcode absent — middleware assigns a synthetic ID from the message control number.

Named patient sample (with barcode):

PID|20|||142|Fatou BA|||F|...
OBR|2|21346159068142|2|...

Barcode 142 links the result set to the Prolab worklist order for that patient.

8. Reference Unit Considerations

Biobase transmits values in conventional units (g/L, mg/L, U/L). Prolab may store in SI units depending on laboratory configuration. The middleware applies conversion factors where needed — for example:

9. Error Handling

10. Deployment Notes

• Middleware runs as a Linux service on the proline host
• Socket must bind before the analyzer is powered on (Biobase connects at boot/result-ready)
• The connection is one-shot per result set — the analyzer connects, sends, disconnects
• No ACK is required by the analyzer; the middleware operates in receive-only mode on this port